unclustered recruitment fails if there are no unclustered bins

[chasemc]

If there are no unclustered bins to start then the Nextflow pipeline fails because of a raise error in the Python script:

https://github.com/KwanLab/Autometa/blob/9b8f00e952094e98cd83027371938730a2933ca4/autometa/binning/unclustered_recruitment.py#L573-L575

[chasemc]

Additional context: I currently have a local branch that pulls isolate genome scaffolds and inputs them for clustering so they should cluster very easily. I'm assuming that these produce no unclustered contigs after the first binning

[chasemc]

Proposed fix (I can do this if it looks ok @WiscEvan @Sidduppal @jason-c-kwan ): Change

 if not prev_num_unclustered:
        raise BinningError("No unclustered contigs are available to recruit!")

to

    if not prev_num_unclustered:
        print("No unclustered contigs are available to recruit!")
        return 

And change the Nextflow outputs to

output.... optional: true  

[evanroyrees]

The return does not necessarily exit the program. We have logging enabled throughout the library so this should be used instead of print.

import sys

if not prev_num_unclustered:
    logger.warning("No unclustered contigs are available to recruit!")
    sys.exit(status=0)

Otherwise, I'm fine with changing this.

[chasemc]

Yeah but sys.exit() also raises an exception so I think that also won't work. The return should just exit main() and be sufficient.

[chasemc]

Oh, status=0, my bad

[chasemc]

Just putting here for posterity:

sys.exit(status=0) gives TypeError: exit() takes no keyword arguments

Changed to: sys.exit(0)

[chasemc]

Tested and

import sys

if not prev_num_unclustered:
    logger.warning("No unclustered contigs are available to recruit!")
    sys.exit(status=0)

works. Adding to next commit

[chasemc]

Fixed by https://github.com/KwanLab/Autometa/pull/157